Title :
Improved frame selector for multi-reference motion estimation
Author :
Kim, Sung Dae ; Sunwoo, Myung Hoon
Author_Institution :
Sch. of Electr. & Comput. Eng., Ajou Univ., Suwon
Abstract :
This paper proposes efficient reference frame selection algorithms to improve rate distortion performance and to reduce additional computation for frame selection. The improved frame selection (IFS) algorithm controls the number of reference frames according to the accuracy of frame selection. In addition, the proposed reference selection pass (RSP) algorithm can simplify early termination conditions by using the results of the previous coded blocks. The simulation results show that the proposed algorithms can eliminate the overhead of early-termination conditions with only 0.05 dB average bjcntegaard delta peak signal to noise ratio degradation.
Keywords :
motion estimation; rate distortion theory; video coding; H.264/AVC; improved frame selection algorithm; multireference motion estimation; rate distortion performance; reference frame selection algorithm; reference selection pass algorithm; signal to noise ratio; Automatic voltage control; Computational complexity; Computational modeling; Degradation; Electronic mail; Image coding; Motion estimation; PSNR; Rate-distortion; Video coding; H.264/AVC; low power design; motion estimation; multi-reference; variable block size;
Conference_Titel :
SoC Design Conference, 2008. ISOCC '08. International
Conference_Location :
Busan
Print_ISBN :
978-1-4244-2598-3
Electronic_ISBN :
978-1-4244-2599-0
DOI :
10.1109/SOCDC.2008.4815614